Purpose: This work addresses the problem of efficient broadcasting of resource allocations descriptors for VoIP traffic in
mobile OFDMA-based wireless systems. Methods: We present the problem of mapping overhead and show that it can be substantially
reduced by using semi-persistent allocations and by taking advantage of the periodicity of VoIP frames, generated by
a multi-phase vocoder. To handle the impact of mobility on the characteristics of the wireless channel we utilize a cross-layer
decision approach that tracks the channel quality and predicts the expected mobile user behavior such that the system may
under-react to channel changes in some cases. Results: We explicitly show how the variability of the wireless channel can be
tracked by using a Markovian model with a set of discrete states. By estimating both state- and transition probabilities of the
multi-state Markovian model, we underline the foundations for a cross-layer decision algorithm that is able to overlook short
transitions in channel states, without changing the modulation and/or the coding schemes. The main advantage of this approach
is the ability to support multiple codecs, or a single codec with different operational modes, both result in different
packet sizes and different periods. Practical relevance: The proposed channel tracking mechanism is simple enough to be
implemented in the base station of any practical OFDMA-based system using the WiMAX or LTE technology